Ken n esaw State Un iver si tyPh .D . A n aly t i cs an d D ata Sci en ce

Data M ining

Research

National ly Recognized

Visual ization

Annual Repor t 20 16 - 20 17

Big Data.

Big Demand.

Without big data analytics, companies are blind and deaf, wandering out onto the Web like deer

on a freeway.

Geoffrey Moore, author of Crossing the Chasm & Inside the Tornado

Tableof

Contents

M e ssa g e f r o m t h e Pr o g r a m D i r e c t o r .......

Pr o g r a m S t a t is t ic s a n d O v e r v ie w ..............

Ac a d e m ic C u r r ic u lu m ......................................

S t u d e n t Te a c h in g ..............................................

Re se a r c h a n d Pu b lic a t io n s...........................

Re se a r c h Po st e r Pr e se n t a t io n s..................

Se ssio n s Pr e se n t a t io n s..................................

An a l y t ic s Ad v iso r y Bo a r d .............................

3

6

12

13

17

18

19

20In t h e N e w s.........................................................

16

Messagefrom

the Program Director

The McKinsey Global Institute has identified that the demand for deep analytical talent will outpace the supply in the United States by almost 200,000 people within three years (source). In response, the White House has launched a "Big Data Research and Development Initiative", to "expand the workforce needed to develop and use Big Data technologies" (source). This theme is echoed by Thomas Davenport?s article "Data Scientist: The Sexiest Job in the 21st Century" (source). These studies - and many others - point to the need for universities to educate and train "Data Scientists" to address this demand. However, few universities in the country currently have pivoted their curricula to develop degree programs in Data Science - defined as the intersection of Statistics, Mathematics and Computer Science. Kennesaw State University's Ph.D. in Analyt ics and Dat a Science is an advanced degree which has been developed to meet the market demand for Data Scientists. This degree will train individuals to translate large,

structured and unstructured, complex datasets into information to improve decision making. This curriculum includes heavy emphasis on programming, data mining, statistical modeling and the mathematical foundations to support these concepts. Importantly, the program also emphasizes communication skills - both oral and written - as well as application and tying results to business and research problems. Because this degree is

a Ph.D., it creates flexibility.

Graduates can either pursue a position in the private or public sector as a "practicing" Data Scientist - where continued demand is expected to greatly outpace the supply - or pursue a position within academia, where they would be uniquely qualified to teach these skills to the next generation.

These studies - and many others - point to the need for universities to educate and train "Data Scientists" to address this demand.

Quotes from the Ph.D. Students

?For me, the power of the Ph.D. in Analytics and Data Science is the leverage of industry partnerships to design a future facing curriculum. The math, statistics and computer science theories and applications taught in the classroom are directly applied to real-world projects in a consulting-like environment. We get to produce applied research that is used in real business applications for clients.?

Jessica Rudd

"Analytics handles the change, continuing change, inevitable change, that is the dominant factor in society today. This great program satisfies the huge demand for research institutes and private sectors. We have collaboration with fortune 500 companies which allows for the real world application of data science."

Yiyun Zhou

"Dr. Priestley does a great job with ?onboarding? meaning she had a welcome reception and orientation for us and was our go-to for contracts and insurance. During the reception, there were collaborators/recruiters from industry, so it is evident that the program is applied. And on our second day on campus, Career Builder came to present to us and recruit for future positions!"

Lauren Staples

"The Program in Analytics and Data Science at Kennesaw State University has enabled me to engage in applied research at the intersection of Statistics and Computer Science. I have appreciated the ability to do both theoretical and applied research."

Bogdan Gadidov

"I appreciate the ability to work on cutting edge research with private sector companies."

Linh Le

Quotes from the Ph.D. Students

Program Statistics and

Overview

Kennesaw State University's Ph.D. in Analyt ics and Dat a Science is an advanced degree which has been developed to meet the market demand for Data Scientists.

This degree will train individuals to translate large, structured and unstructured, complex datasets into information to improve decision making.

This curriculum includes heavy emphasis on programming, data mining, statistical modeling and the mathematical foundations to support these concepts. Importantly, the program also emphasizes communication skills - both oral and written - as well as application and tying results to business and research problems. Because this degree is a Ph.D., it creates flexibility. Graduates can either pursue a position in the private or public sector as a "practicing" Data Scientist - where continued demand is expected to greatly outpace the supply - or pursue a position within academia, where they would be uniquely qualified to teach these skills to the next generation.

The program is managed in cohorts ? students are only accepted to begin one time a year and students take the majority of their courses together as a group. The intention of a cohort approach is that the students form study groups, engage in research and applied projects together and generally

form a community of support, thereby increasing the likelihood of program completion. Informally the students have commented that the program cohort approach facilitates a depth and breadth of community and relationship that they had not anticipated, but greatly value.

After its initial approval in February 2015, the program received over 35 applications for the first cohort of five funded positions to begin August 2015. The ratio of seven applications for every one funded position is particularly

impressive given that the program had very litt le time for advertising and no formal advertising budget. The Center for Statistics and Analytical Research (CSAR) funded all initial advertising (and continues to provide funding) for the program. The program was marketed locally through interstate billboards and nationally (internationally) through datasciencecentral.com and icrunchdata.com, as well as through a dedicated university website ? graduate.kennesaw.edu/datascience .

...this degree is a Ph.D., it creates flexibility. Graduates can either pursue a position in the private or public sector as a "practicing" Data Scientist...

Coh or t 1 Coh or t 2 Coh or t 3

Began Fal l 20 15 Fal l 20 16 Fal l 20 17

Stu den ts 6 6 6

U.S.Ci t i zen s 4 3 3

GA Residen ts 3 1 2

Cou n tr i es of O r i gi n

Ghana, M oldova, Vietnam , Ch ina

Ch ina (3) and I nd ia

Ind ia, Bangladesh , Ch ina

Yr s Sin ce Un der gr adu ate

10 10 16

A ge Ran ge 26-56 26-45 27-56

Per cen t Fem ale 17 50 50

CLASS SIZE

A SUMMARY OF THE FIRST THREE STUDENT COHORTS

A SUMMARY OF THE FIRST THREE STUDENT COHORTS THE FIRST COHORT RECEIVED OVER 35 APPLICATIONS FOR FIVE FUNDED POSITIONS.

CO H O RT 1- BEGA N FA L L 20 15

Stu den t BS an d M S D egr ee BS an d M S I n st i tu t i on

Baidoo, Edw in B.S. App l ied M athem atics, M .S. Com putat ional Finance

CUN Y Brook lyn , Rutger s Un iver si ty

Bucium as, Ser giu B.S. Com puter Science, M .S. I n for m at ion Technology

Academ y of Econom ic Stud ies (M oldova), Souther n Poly techn ic Un iver si ty

Gad idov, Bogdan B.S. I ndustr ial and System s Engineer ing, M .S. Stat ist i cs

Geor gia Tech , Kennesaw State

H ao, Jie B.S. Stat ist i cs, M .S. Com putat ional M athem atics

N or th Ch ina Un iver si ty, East Tennessee State Un iver si ty

L e, L inh B.S. N etwor k engineer ing, M .S. I n for m at ion System s

H anoi Un iver si ty (Vietnam ), M ar shal l Un iver si ty

Vander heyden , Rober t (Funded th rough IBM )

B.S. M athem atics, M .S. M athem atics San D iego State Un iver si ty, V i r gin ia Tech

THE SECOND COHORT RECEIVED OVER 55 APPLICATIONS FOR FIVE FUNDED POSITIONS.

As of December 2016, the program has received over 25 applications for the third cohort to begin August 2017.

CO H O RT 2- BEGA N FA L L 20 16

Stu den t BS an d M S D egr ee BS an d M S I n st i tu t i on

H ebbar, Shashank B.S. M echan ical Engineer ing, M .S. I ndustr ial Engineer ing

Anna Un iver si ty (I nd ia), N or th Carol ina State Un iver si ty

O nwi , M ichael B.S. Aerospace Engineer ing, M BA

Geor gia Tech , Un iver si ty of I nd iana

Rudd, Jessica B.S. An th ropology, M .S. Ep idem iology

Em or y Un iver si ty

W ang, Yan B.S. in Phar m acology, M .S. in Stat ist i cs

Ch ina Phar m aceut ical Un iver si ty, Un iver si ty of Geor gia

Z hang, L i l i B.S. I n for m at ion System s, M .S. I ndustr ial Engineer ing

Cen tr al South Un iver si ty (Ch ina), Un iver si ty of Tennessee

Yiu in , Z hou B.S. Com puter Science, M BA

Chongqing Un iver si ty (Ch ina), Geor gia State Un iver si ty

THE THIRD COHORT RECEIVED OVER 80 APPLICATIONS FOR FIVE FUNDED POSITIONS.

CO H O RT 3- BEGA N FA L L 20 17

Stu den t BS an d M S D egr ee BS an d M S I n st i tu t i on

Andy H enshaw B.S. Electr i cal Engineer ing, M .S. Electr i cal Engineer ing

Geor gia Tech , Geor gia Tech

D ianna Spence B.S. M athem atics, M .S. Com puter Science, Ph .D . M athem atics Educat ion

W i l l iam and M ar y, Geor gia State Un iver si ty, Em or y Un iver si ty

San joosh Akk inen i B.S. Com puter Science, M .S. Com puter Science

Jawahar lal N eh r u Technological Un iver si ty, KSU

M oham m ad M asum

B.S. M athem atics, M .S. M athem atical Sciences

Un iver si ty of D haka, East Tennessee State Un iver si ty

L auren Stap les B.S. Biom edical Engineer ing, M .S. Analy t i cs

Clem son , Un iver si ty of N ew H am psh i re

L iyuan L iu B.S. M athem atics, M .S. Analy t i cs

N an jing Un iver si ty of I n for m at ion Science and Technology, Geor gia State Un iver si ty

Academic Curriculum

The Program curriculum is continuously under review to ensure that the program (1) utilizes the full range of university resources available and (2) is well positioned to place students in both academic and private sector positions. To ensure that the curriculum is fully connected across the Departments of Mathematics, Statistics and Analytical Sciences and Computer Science, representatives from each of these departments have been asked to participate in an ?academic curriculum committee? for the program. This committee meets at least once every semester to review and continuously integrate the courses across departments.

The committee includes the following faculty:

Teaching is an important component of the overall Ph.D. curriculum and is critical to its success. This is true for two core reasons. First, the three supporting curriculum departments ? Statistics and Analytical Sciences, Mathematics, and Computer Science ? are three of the most in demand departments at the university in terms of faculty resources. The Ph.D. program has coordinated with these three department chairs to integrate with advanced courses from the respective masters programs where possible (reducing the need for Ph.D.-only courses with small enrollment numbers). As the Ph.D. program uses faculty resources from these departments, the program returns Ph.D. students as resources to teach lower courses in these respective departments. In other words, the program is ?taking from the top of the faculty stack and replacing at the bottom of the faculty stack.?

Second, all students in the program must teach at least one semester ? even if they do not intend to pursue an academic position after graduation. While a proven record of teaching is an important credential for potential academics, the latent skills associated with teaching (e.g., organization of material, public speaking, critical evaluation of material submission) are also recognized as having value outside academia.

Student Teaching

Most of the Ph.D. students have advanced degrees aligned with mathematics, statistics and engineering. As a result, many have a sufficient number of hours in mathematics and statistics to meet SACS accreditation requirements. In addition all Ph.D. students are required to take the GRAD 9100 course offered through the Graduate College during their first year of study. In this course, students learn how to develop and prepare materials for instruction, how to handle a classroom and how to effectively evaluate student work.

Fall semester 2016 was the first semester that students formally taught courses at KSU. Most of the students who taught were in their third semester of the Ph.D. program. This appears to be an appropriate point in the program for students to begin teaching ? they are sufficiently established in their own program and would have been exposed to the material in the respective MS and undergraduate programs through tutoring and support by working in the Research Center (see below).

Course Course Tit le

MATH 1107 Elementary Statistics

MATH 1160 Elementary Applied Calculus

STAT 3010 Computer Applications of Statistics

STAT 4030 Programing in R

STAT 8030 Programing in R

STAT 4210 Applied Regression Analysis

Boardof

Directors

Research & Publications

Although the Ph.D. program in Analytics and Data Science is an applied program, it has requirements for all students to engage in research and peer-reviewed publication. While peer-review journal publication is held as the standard for dissemination of research, students are encouraged to also engage in all forms of dissemination of research results as they develop their skills and establish their research agendas. To that end, students are encouraged to engage in research poster presentations, conference session presentations (peer-reviewed), as well as journal publications. In addition, it should be noted that in these nascent stages in the life of the Ph.D. program, supporting students in their presentations at regional and national (international) conferences is an important form of advertising and promotion for the program ? many masters students and potential hiring institutions learn about the Ph.D. in Analytics and Data Science at Kennesaw State University through students? presence at national conferences.

Boardof

Directors

Research Poster Presentations

Lili Zhang, Ying Xie, and Guoliang Liu. (2017) A Sentiment-Change-Driven Event Discovery System. Proceedings of WI ?17, Leipzig, Germany, August 23-26, 2017, 7 pages. DOI: 10.1145/3106426.3109038

O?Hagan JJ, Carias C, Rudd JM, (2016) Estimation of Severe Middle East Respiratory Syndrome Cases in the Middle East, 2012?2016. Emerging Infectious Diseases. 2016; 22(10):1797-1799. doi:10.3201/eid2210.151121.

Ying Xie, Linh Le, Jie Hao, (2016) Unsupervised Deep Kernel for High Dimensional Data, Proceeding of the 30th IEEE International Joint Conference on Neural Networks.

Linh Le, Jie Hao, Ying Xie, Jennifer Priestley, (2016) Deep Kernel: Learning the Kernel Function from Data Using Deep Neural Network, Proceedings of the Third IEEE/ACM International Conference on Big Data Computing, Applications and Technologies.

Ying Xie, Chenna Pooja, Linh Le, (2016) Visualization of High Dimensional Data in a Three Dimensional Space, In Proceedings of the Third IEEE/ACM International Conference on Big Data Computing, Applications and Technologies.

Linh Le, Jennifer Priestley, (2017) Using the OPTGRAPH Procedure: Transformation of Transactional Data into Graph for Cluster Analysis, Proceedings of SAS Global Forum 2017.

Gadidov B., Priestley J.L. (2017) Does Yelp Matter? Analyzing (And Guide to Using) Ratings for a Quick Serve Restaurant Chain. Srinivasan S. (eds) Guide to Big Data Applications. Studies in Big Data, vol 26. Springer.

Robert J. McGrath, Jennifer Lewis Priestley, Yiyun Zhou, Patrick J. Culligan, (2017) Do Physicians Ratings Matter? Journal of Medical Internet Research (currently in review).

Publicat ion Awards

Linh Le. Poster winner in SAS Analytics Experience 2016 Conference. Poster tit le: Utilizing Graph Theory in Analyzing Purchase Patterns with PROC OPTGRAPH.

Linh Le. First Place in the HPCC Systems 2016 Poster Presentation Competition ? HPCC Systems Engineering Summit in 2016. Poster tit le: Deep Kernel: Learning the

Kernel Function from the Data.

Cit ed Grey Lit erat ure (in KSU Digit al Com m ons):

An Analysis of Accuracy using Logistic Regression and Time Series (2016), Edwin Baidoo and Jennifer L. Priestley.

A Comparison of Machine Learning Techniques and Logistic Regression Method for the Prediction of Past-Due Amount (2016), Jie Hao and Jennifer L. Priestley.

Application of Isotonic Regression in Predicting Business Risk Scores (2016), Linh T. Le and Jennifer L. Priestley.

Application of Support Vector Machine Modeling and Graph Theory Metrics for Disease Classification (2017), Jessica M. Rudd.

A Comparison of Decision Tree with Logistic Regression Model for Prediction of Worst Non-Financial Payment Status in Commercial Credit (2017), Jessica M. Rudd MPH, GStat and Jennifer L. Priestley. Logistic Ensemble Models (2017), Bob Vanderheyden and Jennifer L. Priestley.

Binary Classification on Past Due of Service Accounts using Logistic Regression and Decision Tree (2017), Yan Wang and Jennifer L. Priestley. Session

Presentations

Analytics Advisory Board

Bi l l Franks, Chief Analytics Officer for Teradata

Wil l Hakes, Managing Director of Date Science - KPMG

Don Hayes, SAS Enterprise Architect and DLL Consulting Inc., a SAS Affiliate Partner

Pet er Maynard, SVP of Enterprise Analytics- Equifax

Jer ry L. Oglesby, Senior Director of Global Academic Programs and Global Certification - SAS

Spencer Taf t , Analytics and Business Intelligence Leader - Cox Enterprises

Mike St efanick , Global Director for Analytics and Technology - EY

Alex Vayner , North American Data Science Practice Leader - CapGemini

IN THE NEWS

1.    The Program received over $250,000 in sponsored research grants from companies including Equifax and GE Power.

2.   Value Colleges ranked KSU #34 on its list of Top 50 Best Value Big Data Graduate Programs.

3.    DataInformed recognized the MSAS as one of the best big data university programs in the U.S.

4.    KSU won first place at the 2017 SAS Global Forum Student Symposium after making up 50% of the finalists.

5. Datanami named Dr. Jennifer Priestley, director of KSU?s Analytics and Data Science Ph.D. program, one of its People to Watch in 2016.

In just the past two years alone, these programs have gained numerous recognitions:

Quotes from the Ph.D. Students

In just the past two years alone, these programs have gained numerous recognitions: "The program provides students with advanced knowledge, tools and skills to analyze big data, collaborate with diverse companies and help students delve further into their research with "real-world" experiences. The program is not only a Ph.D. program, but it is also like "a big family" in which everyone gets close during their studies. The knowledgeable professors, supportive faculty and professional cohort make this program fantastic. I can learn many things from these diverse friendships."  

Liyuan Liu

Quotes from the Ph.D. Students

"It is a pleasure for me to be a part of this Ph.D. program. In this program, we have a fabulous program director who always cares about everyone and gives her best to support us. It provides more opportunities to pursue whatever you are interested in, academia or industrials. Strong relationships with private sector companies allow students to contribute solutions to the real world problems. Moreover, I am able to collaborate with many knowledgeable and supportive professors or researchers, which makes me feel as warm as being home. There is no doubt that this program will be endorsed as one of best graduate data science programs."   

Jie Hao

"The goal is to turn data into information, and information into insight." ? Carly Fiorina, former chief executive of Hewlett-Packard Company or HP.

Kennesaw State University585 Cobb Ave NW, MD 0112Kennesaw Hall, Room 3422

Kennesaw, Georgia 30144-5591(p): 470-578-6107

graduate.kennesaw.edu/datascience